The continuing miniaturization of electronic devices is driving the design of interconnects in the direction of finer pitch surface tracks, smaller diameter through-holes and vias, and thicker workpieces to provide increased circuit densities (Paunovic, M. and M. Schlesinger, 2000)1. This trend has significant implications for the electronics industry which must ensure that the metal electrodeposition process meets the functionality and quality requirements of these advanced workpiece device designs. These workpieces include printed circuit boards, chip scale packages, wafer level packages, printed wiring boards, high density interconnect printed wiring boards, high density interconnect printed circuit boards and the like.
For economies of production, the range of approximate dimensions of workpieces is typically 6 inch by 6 inch, 10 inch by 18 inch, 18 inch by 24 inch, 2 meters by 2 meters, 5 meters by 5 meters, and 200 millimeters and 300 millimeters in diameter. However, these range of dimensions are not unique and are not limiting to the need for controlling the hydrodynamics in a plating cell to facilitate uniform deposition across a workpiece.
As the hole and via diameter decrease, the workpiece thickness increases, and the workpiece dimension increases, the most notable challenge for the quality of metal electrodeposits is the avoidance of non-uniform copper thickness distribution over board surfaces and within through-holes, i.e. the challenge of leveling or throwing power, which can adversely affect the performance of the finished printed wiring board interconnect (Paunovic, M. and M. Schlesinger, 1998)2, (Ward, M., D. R. Gabe and J. N. Crosby, 1999a)3.
A number of operating parameters and plating cell attributes influence the uniformity of copper deposition onto a workpiece. This invention concentrates on the influence of electrochemical cell configuration on the uniformity of copper deposition on the board surface, in particular, the influence of cell configuration on solution hydrodynamics, and the ability to generate uniform flow of electrolyte across the surface of the board during the plating operation. FIG. 1 shows a plating cell (100) which contains a workpiece (102). Although only one workpiece is shown in this and subsequent drawings, one skilled in the art understands that in actual practice a plurality of workpieces may be contained in the plating cell. For ease of description, the term workpiece is understood to encompass one or more workpieces. The workpiece (102) in prior art FIGS. 2-3, 5, and 7-9 is presented as a generally flat panel having at least one generally flat surface for electroplating. Arrows (104) indicate the desired uniform flow of electrolyte across the entire surface of the workpiece (102).
FIG. 2 shows a conventional workpiece (102—shown in a side-view relative to its appearance in FIG. 1) plating operation, in which flow of electrolyte is achieved by air sparging. Air bubbles (106) are created in the electrolyte by blowing air through pipes (108) which have holes in them. These pipes are positioned on the bottom of the plating cell (100) beneath the workpiece (102). The number of pipes (108) is not limited. The movement of air bubbles (106) from the bottom to the top of the plating cell (100) creates solution movement, as indicated by the arrows (104). However, air sparging can create problems in the plating operation:                the oxygen can oxidize components of the electrolyte,        the oxygen can oxidize features and circuit patterns on the workpiece,        air bubbles (106) may become trapped in features in the workpiece (102), creating areas where copper cannot be deposited,        this method can generate low solution movement rates, which can result in burning of the workpiece (102) at high current densities, and        as the air bubbles progress towards the top of the cell they grow in size and can create a non-uniform solution environment from the bottom of the workpiece to the top.        
To avoid the problems associated with air sparging, eductors are being tested for use in plating cells designed for workpieces. Eductors are nozzles which utilize venturi effects to provide up to five times the solution flow velocity output of the pump which feeds the eductors. Eductors may be obtained commercially from a number of sources; one such eductor is marketed under the name Serductor™ (Serductor™ is a trademark of Serfilco, Northbrook, Ill.)4.
One configuration of a prior art plating cell is shown in FIG. 3. The plating cell (100) contains a workpiece (102) which hangs on a rack (110). Anodes (112) are positioned on either side of the workpiece (102) and hang from rails (114). The workpiece (102) serves as the cathode. Eductors (116) are positioned behind the anodes (112) horizontally opposite (perpendicular to) the surface of the workpiece (102) (Weber, A., 2003)5. Fluid flow is directed (shown by the arrows (104)) from the eductors (116) between the anodes (112) to the surface of the workpiece (102). This type of eductor arrangement leads to impinging fluid flow whereby the solution flow velocity is directed toward the workpiece. Solution flow velocity is accomplished through the anodes by openings or spaces in the anodes.
However, as shown in FIG. 4, the use of eductors (116) can lead to a variation in solution flow velocity across the workpiece (102) (Chin, D-T. and C-H. Tsang, 1978), (Hsueh, K-L. and D-T. Chin, 1986a), (Hsueh, K-L. and D-T. Chin, 1986). Fluid flows from the eductor (116) to the impingement point (118) on the surface of the workpiece (102). The fluid flow profile (120) and jet centerline (122) are shown. The flow from the eductor (116) is directly perpendicular to the surface of the workpiece (102). In region I, referred to as the potential core region, the flow from the eductor (116) mixes with the surrounding electrolyte. In region II, referred to as the established flow region, the velocity profile (124) is well established, and the solution flow velocity decreases as a function of distance from the eductor (116). In region III, referred to as the stagnation region, the velocity decreases to almost zero, and the boundary layer thickness is relatively independent of the radial position near the impingement point (118) and centerline (122). In region IV, referred to as the wall jet region, the radial velocity decreases and the boundary layer thickness increases, as a function of distance radially outward from the impingement point (118). These variations in solution flow velocity, termed the glancing effect, within regions III and IV contribute to variations in the thickness of copper deposited on the surface of the workpiece (102).
Efforts to improve the uniformity of flow under the impinging eductor flow configuration have included movement of the workpiece (102) while maintaining the same distance between the workpiece and the eductor (116). while the workpiece movement has generally been reported as left and right, the workpiece movement could conceivably be up and down or even at an angle while maintaining the same distance relative to the eductor. The goal of such movement is to produce a time-averaged uniform boundary layer across the workpiece (102). Such movement, particularly left and right movement is termed knife edge agitation by those skilled in the art. However, knife edge agitation still can result in non-uniformity of the deposited copper and adds complexity to plating cell design. Furthermore, incorporation of knife edge movement in existing workpiece plating lines is difficult and costly.
An alternative prior art configuration shown in FIG. 5 positions the eductors (116) below and off to either side of the workpiece, pointing obliquely at the workpiece surface (102) (Ward, M., D. R. Gabe, and J. N. Crosby, 1998)9, (Ward, M., D. R. Gabe, and J. N. Crosby, 1999b)10.
However, as shown in FIG. 6, the use of angled eductors (116) can lead to a variation in solution flow velocity across the workpiece (102) (Chin, D-T., and M. Agarwal, 1991)11. Fluid flows from the eductor (116) to the impingement point (118) on the surface of the workpiece (102). The fluid flow profile (120) and jet centerline (122) are shown. The flow from the eductor (116) is at an oblique angle to the surface of the workpiece (102). In region I, the potential core region, the flow from the eductor (116) mixes with the surrounding electrolyte. In region II, the established flow region, the velocity profile (124) is well established, and the solution flow velocity decreases as a function of distance from the eductor (116). In region III, the stagnation region, the velocity decreases to almost zero, and the boundary layer thickness is relatively independent of the radial position near the impingement point (118) and centerline (122). In this case, the stagnation point is shifted from the jet centerline. In regions IV and V, the wall jet regions, the velocity decreases and the boundary layer thickness increases, as a function of distance radially outward from the impingement point (118). Furthermore, the solution velocity and boundary layer thickness in region IV is different from that in region V. The glancing effect produces variations in solution flow velocity within regions III, IV, and V, and contributes to variations in the thickness of copper deposited on the surface of the workpiece (102).
An alternative configuration shown in FIG. 7 positions the eductors (116) below and off to either side of the workpiece, pointing obliquely across the workpiece (102) (Weber, A., 2003)5. The eductors (116) on one side of the workpiece (102) are pointed in one direction, and in the opposite direction on the other side (not shown in FIG. 7) of the workpiece (102). This is intended to create a swirling solution movement around the workpiece (102). However, the glancing effect described above applies in this case, leading to non-uniform flow of solution across the workpiece (102).
An alternative configuration shown in FIG. 8 positions the eductors (116) directly below the workpiece (102), pointing directly up so that solution moves past the surface of the workpiece (102) (Weber, A., 2003)5 (Carano, M., 2003)12. Again, the glancing effect described above applies in this case, due to mixing of the flow profiles from the multiple eductors (116) positioned below the workpiece (102). This contributes to non-uniform flow of solution across the workpiece (102).
An alternative configuration shown in FIG. 9 positions the eductors (116) directly below and off to either side of the workpiece (102), pointing directly up so that solution moves past the surface of the workpiece (102) (Carano, M., 2003)12. The glancing effect described above applies in this case, contributing to non-uniform flow of solution across the workpiece (102).
Accordingly, a need exists for a method and apparatus which controls the hydrodynamics within a plating cell (100), to facilitate uniform distribution of metal onto a workpiece (102). This invention concentrates on the influence of cell configuration on the uniformity of deposition across the surface of the workpiece (102) as reflected in a low coefficient of variability.